Skip to main content

Table 3 Summary of difference detection results by combining different statistical tests.

From: Using a spike-in experiment to evaluate analysis of LC-MS data

 

msInspect

MZmine 2

Progenesis

XCMS

Total number of features detected

31168 (12)

12271 (12)

9267 (9)

21486 (13)

Number of features used for statistical analysis

6525 (9)

12092 (9)

8415 (9)

8703 (10)

t-test (q < 0.05)

4824 (9)

3505 (7)

4465 (9)

1896 (7)

Wilcoxon rank-sum test (p < 0.05)

603 (8)

1318 (6)

1584 (8)

812 (8)

t-test (q < 0.05) + Wilcoxon (p < 0.05)

603 (8)

967 (6)

1379 (8)

672 (7)

t-test (q < 0.05) + FC (> 10)

2099 (9)

539 (7)

467 (8)

66 (7)

t-test (q < 0.05) + Wilcoxon (p < 0.05) + FC (> 10)

388 (8)

323 (6)

238 (7)

55 (7)

  1. The total number of features detected and those satisfying different criteria, i.e., t-test with multiple hypothesis testing (q-value < 0.05), Wilcoxon sum-rank test (p-value < 0.05), and fold change (FC > 10). The plus symbol denotes the combination of different criteria. Only features present in at least two replicates in each group were used for statistical analysis. The numbers in parenthesis indicate the number of MassPrep features identified in each category. We found six features (Features 1, 2, 3a, 3b, 4, and 5b) identified as differentially abundant by each of the four software tools in each category